3442
J. Org. Chem. 1991,56,3442-3445
of NC13 on la-c(I) at room temperature.18 Moreover, the isomerization reactions are strongly dependent on the nature of the X1and X2 substituents and the polarity of the solvent. With the 1,2dichloroethaneas solvent, the isomerization of lb is the quickest reaction (tl12 = 6 min at 60 “C) whereas lc is the slowest (til? = 180 min at 70 “C). For compounds la-c, the isomerization always takes place with 1,2-dichloroethaneas solvent while it is never observed in hexane. Finally, only compound 1b undergoes the isomerization reaction in benzene or toluene solution. The rate constant of the interconversiondepends on the polarity of the solventa. For the reaction lb(1) lb(II), tl12 = 20 min at 60 “C in the benzene whereas tl12= 6 min in the dichloroethane at the same temperature. However, in the most polar solvent used (nitrobenzene),compounds la-c(I) do not undergo the expected isomerisations but one ionization reaction takes place which leads to the formation of phosphonium phosphaalkenes 3a-c. The calculated dipolar moment for the conformation 1 and I1 of HP=CHPH, are respectively0.76and 1.25 Debye. This
-
(18) Gouygou, M.;Tachon, C.; Etemad-Moghadam, G.; Koenig, M. Tetrahedron Lett. 1989,90,7411.
-
result is consistent with the influence of polar solvent on the isomerization reactions I 11. The fact, that the diphoaphapropenes substituted on the phosphorus by the TSi group are thermodynamically stable in the syn conformation I1 is probably the consequence of the difference in steric bulkiness and of the electronic effect of Tsi group which, in this case, reduces the stabilizing np-w*F-c interaction.
Conclusion The 1,3-diphosphapropenes exist in the form of two stable rotamers; their existence predicted by a theoretical study is verified experimentally using 31PNMR method. The structural deformations of these two rotamers are connected to differentiate interactions between the P = C system and the phosphino group. These structural changes manifested through different NMR parameters would result in very different reactivities for these isolated systems. Acknowledgment. We thank the Direction Chimie Du CNRS for the calculation time dotation on the VP 200 computer of the CIRCE and the GRECO “Basse Coordinence”.
Notes Bromine-Catalyzed Configuration Isomerization in Bicyclic [3.2.1] Systems Osman Cakmak, Yavuz Tagkesenligil, and Metin Balci* Department of Chemistry, Ataturk University, Faculty of Science, Erzurum, Turkey Received March 9, 1990
The study of ground state and photochemical solvolysis and Wagner-Meerwein rearrangements in certain bridge-ring systems containing aromatic groups and nucleofugal groups has been of considerable interest.’s2 Study of these compounds has shown that there are stereoand regiochemical requirements for both ionization and subsequent rearrangements. In this paper, we describe a new system where similar rearrangements have been observed. We report also an unusual configurational isomerization of endo-2,3-dibromo-6,7-benzobicyclo[3.2.1]octa-3,g-diene (1) in the presence of bromine, which provides evidence for radical rearrangement.
Results and Discussion The starting material 23was synthesized by addition of dibromocarbene to the readily available benzonorbornadiene,’ as reported in the literature. The reaction of 2 with bromines under radical conditions provided 3 in (1) (a) Cristol, S. J.; St”, R. M.J. Am. Chem. SOC.1979,102,5707. (b) Cristol, 5. J.; Dickenson, W.A.; Stanko, M.K. J. Am. Chem. SOC. 1989, 106, 1218. (c) Cristol, S. J.; Seapy, D. G.;Aelling, 0. E. J. Am. Chem. SOC.1989,106,7345. (2) Tanida, H.; Ton, K.; Kitahonoki, K. J. Am. Chem. SOC.1967,89, en, n
JWl1.
(3) Kitahonoki, K.; Takano, Y.;Matsuura, A.; Kotera, K. Tetrahedron 1969,26,335. (4) Mich, T. F.; Nienhouse, E. J.; Farina, T. E.; Tufariello, J. J. J. Chem. Educ. 1968,45, 272.
Scheme I
s & /
/
8r
Br
1
t
LIAIH4
2
Br
1.
a yield of 65% besides six other products.s Treatment of 3 with 1mol of sodium methoxide gave the tribromide 4 (Scheme I). The structure of 4 and the exo position of the bromine was determined by means of ‘H and ‘9c NMR spectra? (5) Treatment of 2 with bromine causes a t first configuration isomerization in 10 min followed by addition of bromine to give 3 as the major (6) akmak,0.; Balci, M.J. Org. Chem. 1989,54, 181. pr”u$ (7) he exo configuration of bromine a t C4 han been established by analysis of the AB system arising from the bridge methylene protons Ha
and HBi. The B part of the AB system show a doublet. There ie no further measurable coupling with the adjacent bridge head protons HI and H6 due to nearly 90° dihedral angles between H b and HI and HI. However, the A part of the AB system is split into triplets of doublets of doublet. The second doublet splitting (‘Ja = 1.5 Hz) originate from the proton on C4 which is in the endo position. In the case of 4Jin the bicyclic systems one speaks of the M or W arrangement. The bonding arrangement of the coupled protons H4 and HBImeets M criterion. The fact that there ie any coupling between Ha and H4 is an indication for the ex0 configuration of the bromine atom at CI.
0022-3263/91/ 1956-3442$02.50/0 0 1991 American Chemical Society
.
Notes
J. Org. Chem., Vol. 56, No.10, 1991 3443 Scheme I1
Scheme I11
&' I
Br
6
& 1
B
I
2
Br
%
& 6
Br
4
+
D
Br
7
Itc
tI &BrI
5
a = hv, b = A, c = FeBr3
LiAIHl reduction of 4 furnished the endo bromide 1. The attack of hydride ion occurs stereospecifically at C2 from the exo face of the molecule, the exo bromine at C4 being displaced according to a SN2' mechanism. Probably, a-substitution sterically retards the normal sN2 mechanism, so that the nucleophile attacks at the y-carbon rather than at the usual position, and an allylic rearrangement takes place. This mechanism is a stereospecific cis process, involving hydride attack from the same side from which bromide leaves. A similar stereochemical course was observed in cyclohexane systems.8 The structure of 1 was ascertained by its 'H NMR spectrum, very similar to that of 2, indicating that the two products are stereoisomers. The interesting feature of these 'H NMR spectra is the AB system arising from the bridge protons H , and Ha in 1 and 2. There is no large chemical shift difference between the resonances of the internal hydrogens Hgi(A6 = 0.1 ppm) in 1 and 2. But H, in 2 resonates a t lower field (6 = 2.6 ppm) compared to the H, (6 = 2.2 ppm) in 1. This fact can be explained on the basis of strong steric repulsion between H , in 2 and neighboring bromine in the exo position. It is well known that interactions related to the van der Waals effect cause a paramagnetic contribution to the shielding constants which results in a shift to lower fieid.9 Pure 1 or 2 were each subjected to direct irradiation with a sun lamp (250 W); a mixture of 1 and 2 in which the later predominated (13:87)was formed. A similar product distribution was also observed by heating of pure samples 1 and 2. The same reactions were conducted in the presence of free radical inhibitors like 2,4,6-tri-tert-buunder the tylphenol and 2,4,6-tri-tert-butylnitrosobenzene ~~~~
(8) Stork, C.; White, W. N. J. Am. Chem. SOC.1956, 78,4609. (9) (a) Tori, K.;Ueyama, M.; Tsuji, T.; Matsamura, H.; Tanida, H.; Iwamura, H.; Kushida, K.; Ninhida, T.; Satoh, 5. Tetrahedron Lett. 1974, 327. (b) Davis, S. C.; Whitman, G. H. J. Chem. SOC., Perkin Trans. 2 1975, 861. ( c ) Crietl, M.; Leininger, H.; Brunn, E. J. Org. Chem. 1982, 47,882. (d) Gheo Mu, M.D.; Olteanu, E. J. Org. Chem. 1987,52,5168. (10) Bala, M.;takmak, 0.; Harmandar, M. Tetrahedron Lett. 1985, 26,5469.
+
0r
endoisomer
Br
D
13
/
&&
/
Br
&-
"'
11
12
same reaction conditions. In these cases no configuration isomerization was observed which strongly supports a radical mechanism for interconversionof the endo and exo bromides (Scheme 11). In both cases, photochemical and thermal reactions, no trace of the [2.2.2] system could be detected. Treatment of endo or exo bromide 1 and 2 with FeBr3 at 100 "C for 2 h gave a mixture of 1,2, and 5 in a ratio of approximately 163846. A similar product ratio was also obtained by treatment of pure sample of 5 with FeBr3 It would appear that thermodynamic control leads largely, but not entirely, to the bicyclo[2.2.2]octadiene system. These results are completely in agreement with those reported by Cristol and Stromla for the dichloro system. Kinetic control of carbenium ion reactions result largely in exo capture of nucleophile to give 2, with smaller amount of endo capture to give 1. The fact that we did not observe any alkyl-shifted product such as 8, indicates
& 0;a. & & c;' Br
Br 8
7
/
9
Br
+: '
Br
10
that the lowest lying (in energy) intermediate in this system is the allylic cation 7 and the bridged phenonium ion 9 is a somewhat higher energy species in the equilibration and there are no kinetically significant ions of the type 10 that lead to 8. Probably these are more strained than either of 7 and 9. A solution of the pure endo bromide 1 in chloroform was treated with bromine in day light for 10 min. The reaction was monitored by 'H NMR spectroscopy. In 10 min the equilibrium mixture consisting from exo and endo dibromides 1 and 2 (14236)was reached (Scheme I). This ratio is in good agreement with the ratio obtained by direct irradiation or heating. The same equilibrium mixture was also obtained by bromination of the exo dibromide 2 with bromine. in 10 min of daylight. In order to shed light on the reaction mechanism of this conversion and to show whether this configuration isomerization takes place a t the C2 carbon atom or it is also accompanied by an allylic rearrangement, we synthesized the deuterated compound 11 by reduction of 4 with LiAIDI. The position and exo configuration of deuterium atom was established on the basis of the 'H NMR spectrum (Scheme 111). Deuterated compound 11 was treated with bromine under the same reaction conditions as reported above. 'H
3444 J. Org. Chem., Vol. 56, No.10, 1991 Scheme IV
1
Notes petroleum ether. The combined organic layers were washed with water (2 x 50 mL), dried, and evaporated to give the crude product. The residue was filtered on a short silica gel column (25g), eluting with petroleum ether, yield 97%. Ftecrystallization from hexane/chloroform (41)afforded colorless crystals: mp 99-100 OC; 'H NMR (360MHz, CDC13,TMS)7.15-7.30 (m, 4 H, aromatic H), 4.75 (dd, 1 H, J = 4.5 and 1.5 Hz, H4), 3.85 (d, 1 H, J = 1 Hz), Hi), 3.80 (dd, 1 H, J = 4.5 and 2.0 Hz, Hs), 2.83 (d, 1 H, B part of AB system, J = 11.5 Hz, Hh), 2.4 (ddt, 1 H,
ApartofABsystem,J=11.5,4.5,1.5Hz,HBi);(90MHz,CDCb, TMS) 149.69,140.87,134.37,127.81,127.60,125.08,121.82,120.15,
56.71,53.05,50.17,39.45;IR (KBr, cm-') 2940,1580,1465,1295, 123O,1160,940, s60; Us mle 390/392/394/396 (M+).Anal. Calcd for C12H&r3: C, 36.68,H, 2.31;Br, 67.46. Found: C, 35.72;H, NMR analysis of the product has revealed the scrambling 2.20;Br, 60.49. of deuterium at C2and C, in a ratio of exactly 1:l. Our (1SR ,2RS ,5RS)-2,3-Dibromo-6,7-beneobicyclo[3.2.1 Iocta2,6-diene (1). To a suspension of 0.58 g (15.26mmol) of LiAlH4 resulta clearly shows that bromine atom at C2changes ita in 100 mL of dry and freshly distilled ether was added dropwise position in 11. The fact that deuterium scrambling is of tribromide 4 in 30 mL of dry ether a solution of 3 g (7.63"01) exactly 1:l implies that the formed intermediate must have during 30 min. The resulting reaction mixture was stirred symmetrical structure from which the products 12 and 13 magnetically at room temperature for 20 h. Wet ether was added can be derived. For this conversion we suggest the folto the resulting reaction mixture (while cooling with an ice bath) lowing mechanism involving the allylic radical 6 (Scheme as long as no reaction was observed. The formed precipitate was IV) dissolved by adding dilute HCl solution. The organic layer was According to this mechanism, a (Brdxmolecule performs washed 3-4 times with water, dried over anhydrous MgS04, a molecule-induced homolysis of 1 to yield radical pairs fdtered, and concentrated to give an oil. The residue was fitered 6, which can explain the scrambling of the deuterium atom. over 20 g of silica gel, eluting with petroleum ether to give endo dibromide as a colorless liquid 1.74 g, yield 73%; 'H NMR (60 Recombination of the allylic radical and bromine radical MHz, CCl,, TMS) 7.38-7.60 (m, 1 H, aromatic), 7.15-7.35 (m, 3 gives 1 and 2 in a ratio of 1486. H,aromatic), 6.71 (d, 1 H, J = 6.8 Hz,HJ, 5.2 (d, 1 H,J = 4.6 Hz, HZ), 3.70 (t, 1 H, J 4.6 Hz, Hi) 3.35 (dd, 1 H, J = 6.8 and Conclusion 3.8 Hz, Hd, 2.45 (dt, B part of AB system, J = 10.4 and 3.8 Hz, Finally, we conclude that bromine catalyses the homoHm),2.20 (d, A part of AB system, J = 10.4 Hz, Hh); IR (neat, lytic cleavage of the carbon-bromine bond and recombiAnal. cm-') 3060,2950,2860,1610,1465,1310,1160,940,860. nation of the radicals formed causes configuration isomCalcd for C12H18r2:C, 45.92; H, 3.21;Br, 50.89 Found C, 45.73; erization. On the basis of the above-mentioned reaction H, 3.06; Br, 50.45. Direct Irradiation of 1 in CC4. A solution of 50 mg (0.15 mechanism we suggeat that all reactions giving cis producta "01) of 1 in 0.5 mL of CCl, was placed into a NMR tube. by addition of bromine to alkenes (like styrene, indene, Deoxygenation was followed by irradiation by a 150-Wprojector etc.) have to be checked on whether cis products are lamp for 5 h. The 'H NMR analysis indicated the formation of formed as suggested by direct syn collapse or they are the equilibrium mixture consisting of 1 and 2 in a ratio of 13:87. formed a8 secondary products according to the aboveProlonged irradiation did not change this ratio. The same mentioned reaction mechanism. equilibrium mixture was also obtained starting from pure 2. When irradiation was carried out in the presence of radical inhibitors, Experimental Section no isomerization was observed. Thermal Reaction of 1 and 2. An Wmg (0.25-"01) sample General Methods. Melting points were determined on a of 1 (or 2)was placed in a NMR tube and heated at 150 OC for Thomas Hoover capillary melting point apparatus and are un2 h. After the mixture was to room temperature, 0.5mL of CCl, cohected. Solvents were concentrated at reduced pressure. Inwas added. The 'H NMR analysis of the mixture indicated the fraredspectra were obtained from solution in 0.1" cells or KBr formation of equilibrium mixture consisting of 1 and 2 in a ratio pellets for solids on a Perkin-Elmer 337 infrared recording of 1486. spectrophotometer. The 'H NMR spectra were recorded on EM 360 Varian, and Brucker W M 400 spectrometers and reported (ISR,4SR,5RS)-2,5-Dibromo-7,8-benzobicyclo[ 2.2.21octa2,7-diene (5). Equilibration of 5,2, and 1. Neat endo dibromide in d units with TMS as the internal standard. Apparent splittings 1 (0.25g, 0.79 mmol) was mixed with a trace of ferric bromide are given in all cases. Mass spectra were recorded on a Finnigan-MAT MS Model 4000 mass spectrometer at an ionizing and heated at 100 O C for 2 h. The product was dissolved in ethyl voltage 70 eV. Analytical thin-layer chromatography (TLC) was ether, wmhed with water, and dried (CaCl,). Evaporation of the solvent left an oil whose 'H NMR spectrum indicated a mixture performed on silica gel So, plates. Column chromawaphy was done on silica gel (60-200 mesh) from Merck Company. of 5: 2, and 1 in a ratio of 46.38:16,respectively. The same reaction (1RS,2RS ,4RS ,5SR)-2,3,3,4-Tetrabromo-6,7-benzo- was also done with pure 2 and 5. In both cases, the same bicyclo[3.2.l]oct-6-ene (3).8To a solution of 2 g (6.37mmol) equilibrium mixture was obtained. of ex0 dibromide 2 in 60 mL of methylene chloride was added (LSR ,2RS ,SRS )-2,3-Dibromo-2-deuterio-6,7-benzodropwise, with stirring and during 3.5 h, a solution of 1.1 g (6.9 bicyclo[3.2.l]octa-3,6-diene (11).To a suspension of 0.2 g (4.76 "01) of bromine in 25 mL of methylene chloride at room tem"01) of LiAlD4 in 30 mL of dry and freshly distilled ether was perature. The reaction flaek was irradiated during the reaction added dropwise to a solution of 1 g (2.54mmol) of tribromide 4 with a 150-W sunlamp. The reaction mixture was allowed to stir in 10 mL of dry ether during 30 min. The resulting reaction l/* h, and the solvent was removed under reduced pressure. The mixture was stirred magnetically at room temperature for 20 h. oily residue was crystallized from CH2C12/petroleumether (31) Same workup was done as described by the synthesis of 1: 'H to give tetrabromide 3 (1970mg, 65%). NMR (60 MHz, CCl,, TMS) 7.40-7.60 (m, 1 H, aromatic), (1SR,aRS,SRS)-2,3,4-Tribromo-6,7-benzobicyc~o[3.2.1]- 7.15-7.35 (m, 3 H, aromatic), 6.71 (d, 1 H), 3.70 (t, 1 H), 2.45 (dt, octa-2,6diene (4). To a solution of 4.74 g of 3 (12mmol) in 70 1 H), 2.20 (d,1 H). mL of dry and freshly distilled tetrahydrofuran was added 0.65 Bromine-CatalyzedConfiguration Isomerization of 1. To g (12 mmol) of sodium methoxide at room temperature while a solution of 500 mg (1.59mmol) of endo dibromide 1 in 5 mL stirring magnetically. The resulting reaction mixture was stirred of chloroform was added a solution of 255 mg (1.59 mmol) of for 2.5 h a t room temperature. The solution was poured into a bromine in 2 mL of CCl, in 2 min. The reaction mixture was mixture of petroleum ether (100mL) and water (100mL). The stirred magnetically at room temperature for 10 min. The solvent layers were separated, and the aqueous phase was extracted with was evaporated, and the 'H NMR analysis of the residue has 6
.
J. Org. Chem. 1991,56,3445-3447 revealed the existence of the equilibrium mixture consisting of 1 and 2 in a ratio of 1486. The same equilibrium mixture was also obtained starting from 2 under the same reaction conditions. Reaction of the Endo Dibromide 1 with Bromine in the Presence of Radical Inhibitors. To a solution of 40 mg (0.12 mmol) of endo dibromide 1 in 0.6 mL of CDCls in a NMR tube was added 16 mg of 2,4,6-tri-tert-butylphenol(0.06 mmol) (or 2,4,6-tri-tert-butylnitrosobenzene)and followed by addition of 20 mg (0.12 mmol) of bromine. After 10 h no configuration isomerization was observed.
3445 Scheme I
-
0 R X
&
7' c\
-co,;cc~l,co
/!-Ox
c-
0
R\
/c=c=c=O
X
II
0
Acknowledgment. We are indebted to Atatiirk University (Grant 1987/40) and Department of Chemistry for financial support of this work and to wish the express their appreciation to Prof. Dr. M. Schlosser (University of Lausanne) for 13C NMR spectra and to both the referees for their very helpful suggestions concerning the radical mechanism. w e t r y NO. 1, 132748-50-6; 2,117404-71-4;3, 117269-29-1; 4, 117269-30-4;5, 117404-74-7;11,132699-64-0;Br2,7726-95-6.
Reactivity of [ (Alky1thio)methylenelketenesin the Gas Phase and Photoelectron Spectra of Thiophen3(2H)-ones1 F. Chuburu, S. Lacombe, and G. Pfister-Guillouzo* Laboratoire de Physico-Chimie Mol6culaire, URA CNRS 474, Avenue de l'Universit6, MOO0 Pau, France
A. Ben Cheik, J. Chuche, and J. C. Pommelet
t
1400
Laborotoire de Chimie Organique Physique, URA CNRS 459, Universitl de Reims, Facult6 des Sciences, Moulin de la Housse, BP 347, 51062 Reims, France Received October 11, 1990
Flash vacuum pyrolysis of substituted Meldrum's acid derivatives is known to give methyleneketenes (alkyl2and alkox?' compounds) (Scheme I). However, alkylthio3ps and alkylaminoS derivatives further react to the corresponding five-membered heterocycles, and thiophen-3(W)-ones or pyrrol-3(W)-ones are generally characterized in solution as the reaction products. The only exception concerns the sterically strained bis(alky1thio) Meldrum's acid derivative 1 [isopropylidene (1,3-dithiolan-2-ylidenelmdonate], for which the gas-phase characterization of [bis(alkylthio)methylene]ketene2 could be performed.' As photoelectron spectroscopy (PES) has proved to be a highly efficient tool for the gas-phase characterization of elusive compounds,' the PES detection of other [(al(1) Application of Photoelectron Spectroscopy to Molecular Properties. 45. Part 44: Lacombe, s.; Pfieter-Guillouzo, G.;Guillemin, J. C.; Denis, J. M. J. Chem. SOC.,Chem. Commun., in prees. (2) Mohmand, 5.;Hirabayaechi, T.; Bock,H. Chem. Ber. 1981, 114, 2609. (3)Ben Cheikh, A.; Dhimane, H.; Pommelet,J. C.; Chuche, J. Tetrahedron Lett. 1988,29, 5919. (4) Chuburu, F.; Lacombe, 5.;Pfieter-Guillouzo, G.; Ben Cheikh, A.; Chuche, J.; Pommelet, J. C. J. Am. Chem. Soc., submitted for publication. (5) Hunter, G.A.; McNab, H. J. Chem. SOC.,Chem. Commun. 1990, 375. (6) McNab, H.; Monahan, L. C. J. Chem. SOC.,Perkin Tiam. 1 1988, 863. 869. ---, --(7) (a) Bock, H.; Solouki, B. Angew. Chem., Znt. Ed. Engl. 1981,20, 427. (b) Bock,H. Phosphorus, Sulfur Silicon Relat. Elem. 1990,49-50, 3. (c) Weatwood, N. P. C. Chem. Soc. Reo. 1989,18,317-343. (d) Schulz, R.; Schweig, A. In Structure and Reactioity; Libman, J. F., Ed.;VCH: New York, 1988; Chapter 8.
in
i4
io
e
in
.1 4 .
io
.
ie
i4
io
e
ie
14
io
. b iw-1
(I
Figure 1. Photoelectron spectra of (a) isopropylidene [l(methy1thio)ethylidenelmalonate(3), (b) the pyrolysis of 3 at 893 K, (c) the difference obtained by digitally subtracting acetone (7). from spectrum b, and (d) pure 5-methylthiophen-3(2H)H)-one ky1thio)methylenelketenes has been attempted and their gas-phase reactivity is described. The flash vacuum pyrolysis of the alkylthio Meldrum's acid derivatives is monitored 'in situ" by photoelectron spectroscopy: the compounds are pyrolyzed in the ionization chamber of the spectrometer (short path pyrolysis: SPP). See ref 8 for a detailed description of the apparatus. When submitted to SPP, alkylthio compounds 3 and 5 begin to split off acetone and C02 at 673 K and the re(8) Vallee, Y.; Ripoll, J. L.; Lacombe, s.; Pfister-Guillouzo,G. J. Chem. Res. Synop. 1990,40, J. Chem. Res., Miniprint 401.
0022-3263/91/1956-3~5$02.50/0 0 1991 American Chemical Society
